NBT BOT

A bot that tells you who it is.

NBT Bot runs website and ecommerce QA for our customers. We sign direct crawler traffic, follow site rules, pace requests, back off when a site tells us to, and do not change identity to get around a block.

Production identity

User-Agent: NBTBot/1.0 (+https://nbt.bot/)
Signature-Agent: "https://nbt.bot"
What it does: Website QA, monitoring and journey testing
Operated by: MyStream Limited
Cloudflare access type: Intermediary
Content use: Reference · no model training

Web Bot Auth live Cloudflare BotBase: not submitted yet

How we behave

Stable identity

Direct crawler requests use the same User-Agent and Web Bot Auth identity. The public verification key is published on this domain.

We read site rules first

robots.txt, Crawl-delay, allowed scope and Content Signals are checked before normal crawling starts.

One traffic budget per site

If several tests target the same site, they share the same site-level limits. Separate runs do not get to pile on independently.

We stop rather than push through

If a site rate-limits, blocks or challenges us, we slow down or stop. We do not swap identity, proxy or transport just to force the request through.

What we test

Depending on the customer’s test, NBT can inspect public pages and assets, robots and sitemap files, products and variants, links, forms, rendered behaviour and customer journeys. Normal discovery is read-only. Any test that changes state has to be explicitly authorised for that customer and run.

Cloudflare registration

For direct crawler traffic, we intend to register NBT in BotBase as an Intermediary with Monitoring & Operations and SEO behaviour. Our content-use level is reference. NBT is not a search-index or model-training crawler.

Verify a request

A User-Agent can be copied. If you need stronger proof, use our verification guide and the public Web Bot Auth key directory.

https://nbt.bot/.well-known/http-message-signatures-directory